Dr A S M Hasan

A new AI-driven research project has launched through a collaboration between Murdoch University and the Lions Eye Institute, aiming to transform how retinal vessels are analysed for better eye health diagnostics. Using deep learning, the initiative focuses on semantic segmentation of arteries and veins within the optic disc and surrounding retina—an area critical to detecting conditions such as glaucoma and diabetic retinopathy.

Leading the project is Dr A S M Hasan, a researcher from Murdoch University, whose earlier work on optic disc segmentation laid the foundation for this next phase.

“By extending our segmentation capabilities to retinal vessels, we’re opening doors to more precise diagnostics and a deeper understanding of ocular conditions,” Dr Hasan said.

Funded by the Lions Eye Institute, the project will run for three to five months and is built using PyTorch, aligned with NVIDIA’s Clara and MONAI frameworks—industry-leading tools in medical AI. Semantic segmentation enables the model to distinguish between arteries and veins, helping clinicians identify subtle changes in vessel structure that may indicate disease.

“We’re not just building models—we’re building tools that could one day help save someone’s sight,” Dr Hasan added.

As AI continues to reshape healthcare, this collaboration between Murdoch University and the Lions Eye Institute stands as a promising step toward smarter, faster, and more accurate eye care.